Merge tag 'drm-fixes-2024-09-13' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fixes from Dave Airlie:
 "Regular fixes pull, the amdgpu JPEG engine fixes are probably the
  biggest, they look to block some register accessing, otherwise there
  are just minor fixes and regression fixes all over.

  nouveau had a regression report going back a few kernels that finally
  got fixed, Not entirely happy with so many changes so late, but they
  all seem quite benign apart from the jpeg one.

  dma-buf/heaps:
   - fix off by one in CMA heap fault handler

  syncobj:
   - fix syncobj leak in drm_syncobj_eventfd_ioctl

  amdgpu:
   - Avoid races between set_drr() functions and dc_state_destruct()
   - Fix regerssion related to zpos
   - Fix regression related to overlay cursor
   - SMU 14.x updates
   - JPEG fixes
   - Silence an UBSAN warning

  amdkfd:
   - Fetch cacheline size from IP discovery

  i915:
   - Prevent a possible int overflow in wq offsets

  xe:
   - Remove a double include
   - Fix null checks and UAF
   - Fix access_ok check in user_fence_create
   - Fix compat IS_DISPLAY_STEP() range
   - OA fix
   - Fixes in show_meminfo

  nouveau:
   - fix GP10x regression on boot

  stm:
   - add COMMON_CLK dep

  rockchip:
   - iommu api change

  tegra:
   - iommu api change"

* tag 'drm-fixes-2024-09-13' of https://gitlab.freedesktop.org/drm/kernel: (25 commits)
  drm/xe/client: add missing bo locking in show_meminfo()
  drm/xe/client: fix deadlock in show_meminfo()
  drm/xe/oa: Enable Xe2+ PES disaggregation
  drm/xe/display: fix compat IS_DISPLAY_STEP() range end
  drm/xe: Fix access_ok check in user_fence_create
  drm/xe: Fix possible UAF in guc_exec_queue_process_msg
  drm/xe: Remove fence check from send_tlb_invalidation
  drm/xe/gt: Remove double include
  drm/amd/display: Add all planes on CRTC to state for overlay cursor
  drm/amdgpu/atomfirmware: Silence UBSAN warning
  drm/amd/amdgpu: apply command submission parser for JPEG v1
  drm/amd/amdgpu: apply command submission parser for JPEG v2+
  drm/amd/pm: fix the pp_dpm_pcie issue on smu v14.0.2/3
  drm/amd/pm: update the features set on smu v14.0.2/3
  drm/amd/display: Do not reset planes based on crtc zpos_changed
  drm/amd/display: Avoid race between dcn35_set_drr() and dc_state_destruct()
  drm/amd/display: Avoid race between dcn10_set_drr() and dc_state_destruct()
  drm/amdkfd: Add cache line size info
  drm/tegra: Use iommu_paging_domain_alloc()
  drm/rockchip: Use iommu_paging_domain_alloc()
  ...

Author: torvalds

drivers/dma-buf/heaps/cma_heap.c

@@ -165,7 +165,7 @@ static vm_fault_t cma_heap_vm_fault(struct vm_fault *vmf)
 	struct vm_area_struct *vma = vmf->vma;
 	struct cma_heap_buffer *buffer = vma->vm_private_data;
 
-	if (vmf->pgoff > buffer->pagecount)
+	if (vmf->pgoff >= buffer->pagecount)
 		return VM_FAULT_SIGBUS;
 
 	return vmf_insert_pfn(vma, vmf->address, page_to_pfn(buffer->pages[vmf->pgoff]));

drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c

@@ -23,6 +23,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_jpeg.h"
+#include "amdgpu_cs.h"
 #include "soc15.h"
 #include "soc15d.h"
 #include "vcn_v1_0.h"
@@ -34,6 +35,9 @@
 static void jpeg_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev);
 static void jpeg_v1_0_set_irq_funcs(struct amdgpu_device *adev);
 static void jpeg_v1_0_ring_begin_use(struct amdgpu_ring *ring);
+static int jpeg_v1_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
+				     struct amdgpu_job *job,
+				     struct amdgpu_ib *ib);
 
 static void jpeg_v1_0_decode_ring_patch_wreg(struct amdgpu_ring *ring, uint32_t *ptr, uint32_t reg_offset, uint32_t val)
 {
@@ -300,7 +304,10 @@ static void jpeg_v1_0_decode_ring_emit_ib(struct amdgpu_ring *ring,
 
 	amdgpu_ring_write(ring,
 		PACKETJ(SOC15_REG_OFFSET(JPEG, 0, mmUVD_LMI_JRBC_IB_VMID), 0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, (vmid | (vmid << 4)));
+	if (ring->funcs->parse_cs)
+		amdgpu_ring_write(ring, 0);
+	else
+		amdgpu_ring_write(ring, (vmid | (vmid << 4)));
 
 	amdgpu_ring_write(ring,
 		PACKETJ(SOC15_REG_OFFSET(JPEG, 0, mmUVD_LMI_JPEG_VMID), 0, 0, PACKETJ_TYPE0));
@@ -554,6 +561,7 @@ static const struct amdgpu_ring_funcs jpeg_v1_0_decode_ring_vm_funcs = {
 	.get_rptr = jpeg_v1_0_decode_ring_get_rptr,
 	.get_wptr = jpeg_v1_0_decode_ring_get_wptr,
 	.set_wptr = jpeg_v1_0_decode_ring_set_wptr,
+	.parse_cs = jpeg_v1_dec_ring_parse_cs,
 	.emit_frame_size =
 		6 + 6 + /* hdp invalidate / flush */
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
@@ -611,3 +619,69 @@ static void jpeg_v1_0_ring_begin_use(struct amdgpu_ring *ring)
 
 	vcn_v1_0_set_pg_for_begin_use(ring, set_clocks);
 }
+
+/**
+ * jpeg_v1_dec_ring_parse_cs - command submission parser
+ *
+ * @parser: Command submission parser context
+ * @job: the job to parse
+ * @ib: the IB to parse
+ *
+ * Parse the command stream, return -EINVAL for invalid packet,
+ * 0 otherwise
+ */
+static int jpeg_v1_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
+				     struct amdgpu_job *job,
+				     struct amdgpu_ib *ib)
+{
+	u32 i, reg, res, cond, type;
+	int ret = 0;
+	struct amdgpu_device *adev = parser->adev;
+
+	for (i = 0; i < ib->length_dw ; i += 2) {
+		reg  = CP_PACKETJ_GET_REG(ib->ptr[i]);
+		res  = CP_PACKETJ_GET_RES(ib->ptr[i]);
+		cond = CP_PACKETJ_GET_COND(ib->ptr[i]);
+		type = CP_PACKETJ_GET_TYPE(ib->ptr[i]);
+
+		if (res || cond != PACKETJ_CONDITION_CHECK0) /* only allow 0 for now */
+			return -EINVAL;
+
+		if (reg >= JPEG_V1_REG_RANGE_START && reg <= JPEG_V1_REG_RANGE_END)
+			continue;
+
+		switch (type) {
+		case PACKETJ_TYPE0:
+			if (reg != JPEG_V1_LMI_JPEG_WRITE_64BIT_BAR_HIGH &&
+			    reg != JPEG_V1_LMI_JPEG_WRITE_64BIT_BAR_LOW &&
+			    reg != JPEG_V1_LMI_JPEG_READ_64BIT_BAR_HIGH &&
+			    reg != JPEG_V1_LMI_JPEG_READ_64BIT_BAR_LOW &&
+			    reg != JPEG_V1_REG_CTX_INDEX &&
+			    reg != JPEG_V1_REG_CTX_DATA) {
+				ret = -EINVAL;
+			}
+			break;
+		case PACKETJ_TYPE1:
+			if (reg != JPEG_V1_REG_CTX_DATA)
+				ret = -EINVAL;
+			break;
+		case PACKETJ_TYPE3:
+			if (reg != JPEG_V1_REG_SOFT_RESET)
+				ret = -EINVAL;
+			break;
+		case PACKETJ_TYPE6:
+			if (ib->ptr[i] != CP_PACKETJ_NOP)
+				ret = -EINVAL;
+			break;
+		default:
+			ret = -EINVAL;
+		}
+
+		if (ret) {
+			dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
+			break;
+		}
+	}
+
+	return ret;
+}

drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.h

@@ -29,4 +29,15 @@ int jpeg_v1_0_sw_init(void *handle);
 void jpeg_v1_0_sw_fini(void *handle);
 void jpeg_v1_0_start(struct amdgpu_device *adev, int mode);
 
+#define JPEG_V1_REG_RANGE_START	0x8000
+#define JPEG_V1_REG_RANGE_END	0x803f
+
+#define JPEG_V1_LMI_JPEG_WRITE_64BIT_BAR_HIGH	0x8238
+#define JPEG_V1_LMI_JPEG_WRITE_64BIT_BAR_LOW	0x8239
+#define JPEG_V1_LMI_JPEG_READ_64BIT_BAR_HIGH	0x825a
+#define JPEG_V1_LMI_JPEG_READ_64BIT_BAR_LOW	0x825b
+#define JPEG_V1_REG_CTX_INDEX			0x8328
+#define JPEG_V1_REG_CTX_DATA			0x8329
+#define JPEG_V1_REG_SOFT_RESET			0x83a0
+
 #endif /*__JPEG_V1_0_H__*/

drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c

@@ -23,6 +23,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_jpeg.h"
+#include "amdgpu_cs.h"
 #include "amdgpu_pm.h"
 #include "soc15.h"
 #include "soc15d.h"
@@ -538,7 +539,11 @@ void jpeg_v2_0_dec_ring_emit_ib(struct amdgpu_ring *ring,
 
 	amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_IB_VMID_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
+
+	if (ring->funcs->parse_cs)
+		amdgpu_ring_write(ring, 0);
+	else
+		amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
 
 	amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JPEG_VMID_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
@@ -764,6 +769,7 @@ static const struct amdgpu_ring_funcs jpeg_v2_0_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v2_0_dec_ring_get_rptr,
 	.get_wptr = jpeg_v2_0_dec_ring_get_wptr,
 	.set_wptr = jpeg_v2_0_dec_ring_set_wptr,
+	.parse_cs = jpeg_v2_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
@@ -810,3 +816,58 @@ const struct amdgpu_ip_block_version jpeg_v2_0_ip_block = {
 		.rev = 0,
 		.funcs = &jpeg_v2_0_ip_funcs,
 };
+
+/**
+ * jpeg_v2_dec_ring_parse_cs - command submission parser
+ *
+ * @parser: Command submission parser context
+ * @job: the job to parse
+ * @ib: the IB to parse
+ *
+ * Parse the command stream, return -EINVAL for invalid packet,
+ * 0 otherwise
+ */
+int jpeg_v2_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
+			      struct amdgpu_job *job,
+			      struct amdgpu_ib *ib)
+{
+	u32 i, reg, res, cond, type;
+	struct amdgpu_device *adev = parser->adev;
+
+	for (i = 0; i < ib->length_dw ; i += 2) {
+		reg  = CP_PACKETJ_GET_REG(ib->ptr[i]);
+		res  = CP_PACKETJ_GET_RES(ib->ptr[i]);
+		cond = CP_PACKETJ_GET_COND(ib->ptr[i]);
+		type = CP_PACKETJ_GET_TYPE(ib->ptr[i]);
+
+		if (res) /* only support 0 at the moment */
+			return -EINVAL;
+
+		switch (type) {
+		case PACKETJ_TYPE0:
+			if (cond != PACKETJ_CONDITION_CHECK0 || reg < JPEG_REG_RANGE_START ||
+			    reg > JPEG_REG_RANGE_END) {
+				dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
+				return -EINVAL;
+			}
+			break;
+		case PACKETJ_TYPE3:
+			if (cond != PACKETJ_CONDITION_CHECK3 || reg < JPEG_REG_RANGE_START ||
+			    reg > JPEG_REG_RANGE_END) {
+				dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
+				return -EINVAL;
+			}
+			break;
+		case PACKETJ_TYPE6:
+			if (ib->ptr[i] == CP_PACKETJ_NOP)
+				continue;
+			dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
+			return -EINVAL;
+		default:
+			dev_err(adev->dev, "Unknown packet type %d !\n", type);
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}

drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.h

@@ -45,6 +45,9 @@
 
 #define JRBC_DEC_EXTERNAL_REG_WRITE_ADDR				0x18000
 
+#define JPEG_REG_RANGE_START						0x4000
+#define JPEG_REG_RANGE_END						0x41c2
+
 void jpeg_v2_0_dec_ring_insert_start(struct amdgpu_ring *ring);
 void jpeg_v2_0_dec_ring_insert_end(struct amdgpu_ring *ring);
 void jpeg_v2_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
@@ -57,6 +60,9 @@ void jpeg_v2_0_dec_ring_emit_vm_flush(struct amdgpu_ring *ring,
 				unsigned vmid, uint64_t pd_addr);
 void jpeg_v2_0_dec_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val);
 void jpeg_v2_0_dec_ring_nop(struct amdgpu_ring *ring, uint32_t count);
+int jpeg_v2_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
+			      struct amdgpu_job *job,
+			      struct amdgpu_ib *ib);
 
 extern const struct amdgpu_ip_block_version jpeg_v2_0_ip_block;
 

drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c

@@ -662,6 +662,7 @@ static const struct amdgpu_ring_funcs jpeg_v2_5_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v2_5_dec_ring_get_rptr,
 	.get_wptr = jpeg_v2_5_dec_ring_get_wptr,
 	.set_wptr = jpeg_v2_5_dec_ring_set_wptr,
+	.parse_cs = jpeg_v2_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
@@ -691,6 +692,7 @@ static const struct amdgpu_ring_funcs jpeg_v2_6_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v2_5_dec_ring_get_rptr,
 	.get_wptr = jpeg_v2_5_dec_ring_get_wptr,
 	.set_wptr = jpeg_v2_5_dec_ring_set_wptr,
+	.parse_cs = jpeg_v2_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +

drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c

@@ -560,6 +560,7 @@ static const struct amdgpu_ring_funcs jpeg_v3_0_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v3_0_dec_ring_get_rptr,
 	.get_wptr = jpeg_v3_0_dec_ring_get_wptr,
 	.set_wptr = jpeg_v3_0_dec_ring_set_wptr,
+	.parse_cs = jpeg_v2_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +

drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c

@@ -727,6 +727,7 @@ static const struct amdgpu_ring_funcs jpeg_v4_0_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v4_0_dec_ring_get_rptr,
 	.get_wptr = jpeg_v4_0_dec_ring_get_wptr,
 	.set_wptr = jpeg_v4_0_dec_ring_set_wptr,
+	.parse_cs = jpeg_v2_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +

drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.h

@@ -32,5 +32,4 @@ enum amdgpu_jpeg_v4_0_sub_block {
 };
 
 extern const struct amdgpu_ip_block_version jpeg_v4_0_ip_block;
-
 #endif /* __JPEG_V4_0_H__ */

drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c

@@ -23,9 +23,9 @@
 
 #include "amdgpu.h"
 #include "amdgpu_jpeg.h"
-#include "amdgpu_cs.h"
 #include "soc15.h"
 #include "soc15d.h"
+#include "jpeg_v2_0.h"
 #include "jpeg_v4_0_3.h"
 #include "mmsch_v4_0_3.h"
 
@@ -1089,7 +1089,7 @@ static const struct amdgpu_ring_funcs jpeg_v4_0_3_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v4_0_3_dec_ring_get_rptr,
 	.get_wptr = jpeg_v4_0_3_dec_ring_get_wptr,
 	.set_wptr = jpeg_v4_0_3_dec_ring_set_wptr,
-	.parse_cs = jpeg_v4_0_3_dec_ring_parse_cs,
+	.parse_cs = jpeg_v2_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
@@ -1254,56 +1254,3 @@ static void jpeg_v4_0_3_set_ras_funcs(struct amdgpu_device *adev)
 {
 	adev->jpeg.ras = &jpeg_v4_0_3_ras;
 }
-
-/**
- * jpeg_v4_0_3_dec_ring_parse_cs - command submission parser
- *
- * @parser: Command submission parser context
- * @job: the job to parse
- * @ib: the IB to parse
- *
- * Parse the command stream, return -EINVAL for invalid packet,
- * 0 otherwise
- */
-int jpeg_v4_0_3_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
-			     struct amdgpu_job *job,
-			     struct amdgpu_ib *ib)
-{
-	uint32_t i, reg, res, cond, type;
-	struct amdgpu_device *adev = parser->adev;
-
-	for (i = 0; i < ib->length_dw ; i += 2) {
-		reg  = CP_PACKETJ_GET_REG(ib->ptr[i]);
-		res  = CP_PACKETJ_GET_RES(ib->ptr[i]);
-		cond = CP_PACKETJ_GET_COND(ib->ptr[i]);
-		type = CP_PACKETJ_GET_TYPE(ib->ptr[i]);
-
-		if (res) /* only support 0 at the moment */
-			return -EINVAL;
-
-		switch (type) {
-		case PACKETJ_TYPE0:
-			if (cond != PACKETJ_CONDITION_CHECK0 || reg < JPEG_REG_RANGE_START || reg > JPEG_REG_RANGE_END) {
-				dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
-				return -EINVAL;
-			}
-			break;
-		case PACKETJ_TYPE3:
-			if (cond != PACKETJ_CONDITION_CHECK3 || reg < JPEG_REG_RANGE_START || reg > JPEG_REG_RANGE_END) {
-				dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
-				return -EINVAL;
-			}
-			break;
-		case PACKETJ_TYPE6:
-			if (ib->ptr[i] == CP_PACKETJ_NOP)
-				continue;
-			dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
-			return -EINVAL;
-		default:
-			dev_err(adev->dev, "Unknown packet type %d !\n", type);
-			return -EINVAL;
-		}
-	}
-
-	return 0;
-}